info before making it public
Browse files
README.md
CHANGED
@@ -1,4 +1,6 @@
|
|
1 |
---
|
|
|
|
|
2 |
license: cc-by-4.0
|
3 |
task_categories:
|
4 |
- token-classification
|
@@ -13,18 +15,52 @@ size_categories:
|
|
13 |
# Function Of Citation in Astrophysics Literature (FOCAL): Dataset and Task
|
14 |
Can you explain why the authors made a given citation?
|
15 |
|
16 |
-
|
17 |
-
Given a sample paragraph and a query citation within, find where and why that citation was made.
|
18 |
-
Manually tagged text fragments from astrophysics papers, provided by the [NASA Astrophysical Data System](https://ui.adsabs.harvard.edu/)
|
19 |
-
|
20 |
-
Examples:
|
21 |
-
`provide an example paragraph with 2 different queries`
|
22 |
|
23 |
## Dataset Description
|
24 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
|
26 |
-
## List
|
27 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
28 |
|
29 |
Maintainer: Felix Grezes (ORCID: 0000-0001-8714-7774)
|
30 |
Data annotator: Tom Allen (ORCID: 0000-0002-5532-4809)
|
|
|
1 |
---
|
2 |
+
annotations_creators:
|
3 |
+
- expert-generated
|
4 |
license: cc-by-4.0
|
5 |
task_categories:
|
6 |
- token-classification
|
|
|
15 |
# Function Of Citation in Astrophysics Literature (FOCAL): Dataset and Task
|
16 |
Can you explain why the authors made a given citation?
|
17 |
|
18 |
+
This dataset was created as a [shared task](https://ui.adsabs.harvard.edu/WIESP/2023/shared_task_1) for [WIESP @ AACL-IJCNLP 2023](https://ui.adsabs.harvard.edu/WIESP/2023/).
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
## Dataset Description
|
21 |
+
Datasets are in JSON Lines format (each line is a json dictionary).
|
22 |
+
|
23 |
+
Each entry consists of a dictionary with the following keys:
|
24 |
+
- `"Identifier"`: unique string to identify the entry
|
25 |
+
- `"Paragraph"`: text string
|
26 |
+
- `"Citation Text"`: list of strings forming the citation (most often a single string, but sometimes the citation text is split up)
|
27 |
+
- `"Citation Start End"`: list of integer pairs denoting where the citation starts and end in `"Paragraph"` (most often a single pair, sometimes the citation text is split up, if so follows the order in `"Citation Text"`)
|
28 |
+
- `"Functions Text"`: list of strings highlighting parts of the paragraph that explain the function of the citation
|
29 |
+
- `"Functions Label"`: list of strings with the label for each text element in `"Functions Text"` (in same order)
|
30 |
+
- `"Functions Start End"`: list of integer pairs denoting where the elements in `"Functions Text"` start and end in `"Paragraph"`(in same order)
|
31 |
+
|
32 |
+
start and end are defined by the character position in the `"Paragraph"` string.
|
33 |
+
|
34 |
+
|
35 |
+
## Instructions for Workshop Participants:
|
36 |
+
How to load the data using the Huggingface library:
|
37 |
+
```python
|
38 |
+
from datasets import load_dataset
|
39 |
+
dataset = load_dataset("adsabs/FOCAL") # DOES NOT LOAD VALIDATION-NO-LABELS
|
40 |
+
```
|
41 |
+
|
42 |
+
How to load the data if you cloned the repository locally:
|
43 |
+
(assuming `./FOCAL-TRAINING.jsonl` is in the current directory, change as needed)
|
44 |
+
- python (as list of dictionaries):
|
45 |
+
```python
|
46 |
+
import json
|
47 |
+
with open("./FOCAL-TRAINING.jsonl", 'r') as f:
|
48 |
+
focal_training_from_json = [json.loads(l) for l in list(f)]
|
49 |
+
```
|
50 |
+
- into Huggingface (as a Huggingface Dataset):
|
51 |
+
```python
|
52 |
+
from datasets import Dataset
|
53 |
+
focal_training_from_json = Dataset.from_json(path_or_paths="./FOCAL-TRAINING.jsonl")
|
54 |
+
```
|
55 |
|
56 |
+
## File List
|
57 |
+
```
|
58 |
+
├── FOCAL-TRAINING.jsonl (2421 samples for training)
|
59 |
+
├── FOCAL-VALIDATION-NO-LABELS.jsonl (606 samples for validation without the labels. Used during the shared task [WIESP-2023](https://ui.adsabs.harvard.edu/WIESP/2023/)
|
60 |
+
├──
|
61 |
+
├── README.MD (this file)
|
62 |
+
└──
|
63 |
+
```
|
64 |
|
65 |
Maintainer: Felix Grezes (ORCID: 0000-0001-8714-7774)
|
66 |
Data annotator: Tom Allen (ORCID: 0000-0002-5532-4809)
|